arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2306.02671cs.CL
收藏

通过分解与约束改进基于语法的序列到序列建模

Improving Grammar-based Sequence-to-Sequence Modeling with Decomposition and Constraints

ShanghaiTech University(上海科技大学);Shanghai Engineering Research Center of Intelligent Vision and Imaging(上海智能视觉影像工程技术研究中心)

Chao Lou, Kewei Tu

AI总结:

针对Neural QCFG推理昂贵的问题,提出两种低秩变体并引入树层次与源覆盖度软约束,在多数实验设置下提升了性能。

摘要:

Neural QCFG是一种基于语法的序列到序列(seq2seq)模型,对层次结构具有很强的归纳偏置。它在可解释性和泛化能力方面表现出色,但存在推理代价高昂的问题。本文研究Neural QCFG的两种低秩变体以加快推理,并在效率与表达能力之间作出不同权衡。此外,利用语法提供的符号接口,我们引入了关于树层次和源覆盖度的两个软约束。我们在多个数据集上进行实验,结果表明所提模型在大多数设置下均优于原始Neural QCFG。

英文摘要:

Neural QCFG is a grammar-based sequence-tosequence (seq2seq) model with strong inductive biases on hierarchical structures. It excels in interpretability and generalization but suffers from expensive inference. In this paper, we study two low-rank variants of Neural QCFG for faster inference with different trade-offs between efficiency and expressiveness. Furthermore, utilizing the symbolic interface provided by the grammar, we introduce two soft constraints over tree hierarchy and source coverage. We experiment with various datasets and find that our models outperform vanilla Neural QCFG in most settings.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑